We conducted this prospective observational study to identify the incidence, risk factors, and outcomes of sepsis in post-craniotomy critical ill patients for two years. We found the incidence of sepsis in our patients was 33.3%. Male sex, age, tumors of the sellar region, postoperative hydrocephalus, higher APACHE II score, and higher SOFA score were independent risk factors for sepsis. In contrast, tumors of the cranial and paraspinal nerves, infratentorial surgery, and higher GCS score were associated with a lower risk of sepsis. Septic patients had significantly higher hospital mortality rates, lower GOS at discharge, prolonged ICU LOS, prolonged hospital LOS, and higher total hospital costs.
Our results suggested that sepsis was common in post-craniotomy patients admitted into ICU. Compared with previous studies, the incidence of sepsis in our study was relatively lower. The difference may be mainly due to the disparity in the patient population. Most previous studies included medical and/ or surgical patients, while post-craniotomy patients were rarely involved [9, 30]. The incidence of sepsis varied among different populations[7, 8, 14, 19, 31]. Even in the same study, the sepsis occurrence rates have significant differences among medical, scheduled, and unscheduled patients[32]. Changes in definitions of sepsis could partly explain the variation of incidence. Most previous studies have defined sepsis as systemic inflammatory response syndrome (SIRS) due to infection[9, 19, 30], while we defined sepsis according to Sepsis-3 criteria. SIRS has proven to be extremely sensitive but poor specificity for sepsis[9, 22] and may overestimate the sepsis incidence. However, the incidence of sepsis in our study was still higher than those in some other studies. In the study of Berger et al.[20], 12.6% of stroke patients developed sepsis. However, they only counted sepsis episodes that occurred within the first 7 days from stroke onset. The ignoring of sepsis episodes that occurred 7 days later might have resulted in a lower sepsis incidence.
It is a pity that a review of the literature did not yield any studies that included similar patients and was comparable with ours. Some studies have described the epidemiology of infections in neurological critical ill patients [5, 33–35]. Like previous studies[5, 34, 35], pneumonia was the most prevalent infection site, and it was the most common cause of sepsis in our patients. The high incidence of pneumonia highlighted the necessity of early mobility, selective oral or digestive decontamination, subglottic secretion drainage, early tracheotomy, et al. The Society for Healthcare Epidemiology of America (SHEA) and the Infectious Diseases Society of America (IDSA) had recommended those measures [36], which are not implemented or poorly implemented in our hospital.
The incidence of CNS infection was also high in our patients. removing unnecessary drainage/ monitoring intracranial tubes[37], antibiotic prophylaxis might be helpful for the prevention of meningitis and surgical site infection after craniotomy[38, 39].
Prophylactic antibiotics were used in more than half of the patients (n = 462). The second-generation cephalosporins were the most commonly used drugs because of the reimbursement policy. However, antibiotic prophylaxis did not reduce the overall incidence of CNS infection in our study. Patients receiving prophylactic antibiotics even had a higher CNS infection rate (29.6% vs 21.2%, p = 0.004). This might be due to the higher risk of infection in those patients receiving prophylactic antibiotic therapy. Furthermore, according to the report of China Antimicrobial Resistance Surveillance System, antibiotic-resistant bacteria were prevalent in hospitals of China[40]. For example, two thirds (75.4%) of the coagulase-negative Staphylococcus isolated in 2019 were resistant to Methicillin. Multidrug-resistant (MDR) bacteria would be more prevalent in ICU than non-ICU settings[41]. Increasing risk of drug-resistant bacterial infection might have diminished the prophylactic effect of antibiotics.
Interestingly, we found that CNS and surgical site infections were less likely to develop sepsis than pneumonia and gastroenteritis. Such observation might be explained by the difference in pathogen distribution and further confirmation in large multi-center cohorts. Whether it is related to the role of the blood-brain barrier is not clear.
Similar to previous studies, we found that male sex[42], older in age[42, 43], higher APACHE II score[44], higher SOFA score[19] and lower GCS score[19] were independent risk factors for sepsis. Supratentorial surgery, tumors of the sellar region and postoperative hydrocephalus were also independent predictors of sepsis acquisition in our post-craniotomy patients, while tumors of the cranial and paraspinal nerves was associated with a lower risk of sepsis. Such information might help clinicians identify patients at high risk of sepsis, and improve their prognosis through close monitoring and active treatment.
The mortality rate of sepsis in our patients was lower than those in general ICU wards[15, 30]. Our patients were much younger and with fewer comorbidities, which might be the most important reasons for the low mortality rate. In our study, sepsis conferred a 1-fold increase in the odds of death and a 2-fold increase in the odds of prolonged ICU stay, suggesting the importance of early diagnosis and timely sepsis treatment, such as early antibiotic therapy and optimal resuscitation[45–47].